Computer generation of cloze tasks still falls short of full automation; most current systems are used by teachers as authoring aids. Improved methods to estimate cloze quality are needed for full automation. We investigated lexical reading difficulty as a novel automatic estimator of cloze quality, to which cooccurrence frequency of words was compared as an alternate estimator. Rather than rel...

The notion of expectancy grammar as a key to understanding the nature of psychologically real processes that underlie language use is introduced by Oller (1979). A central issue in this notion is that expectancy generating systems are constructed and modified in the course of language acquisition. Thus, one of the characteristics of language proficiency is that it consists of such an expectancy...

Test method facet is one of the factors which can have an influence on the test takers’ performance. The purpose of the current study was to investigate the effects of two different response types, multiple-choice cloze and multiple-choice test, on the pre-intermediate and intermediate test takers’ reading comprehension performance. To this end, 40 pre-intermediate and intermediate learners par...

We present evidence for a monitoring process in language perception at the word level, reflected by a P600. This P600 is triggered when a conflict evolves because the brain encounters an unexpected linguistic item when another item is highly expected. To resolve this conflict between representations, the brain monitors the input to check for possible processing errors. A P600 was hypothesized t...

Cloze-style queries are representative problems in reading comprehension. Over the past few months, we have seen much progress that utilizing neural network approach to solve Cloze-style questions. In this paper, we present a novel model called attention-over-attention reader for the Cloze-style reading comprehension task. Our model aims to place another attention mechanism over the document-le...

We describe the automated generation and use of 69,326 comprehension cloze questions and 5,668 vocabulary matching questions in the 2001-2002 version of Project LISTEN’s Reading Tutor used by 364 students in grades 1-9 at seven schools. To validate our methods, we used students’ performance on these multiple-choice questions to predict their scores on the Woodcock Reading Mastery Test. A model ...

Can automatically generated questions scaffold reading comprehension? We automated three kinds of multiple-choice questions in children’s assisted reading: 1. Whquestions: ask a generically worded What/Where/When question. 2. Sentence prediction: ask which of three sentences belongs next. 3. Cloze: ask which of four words best fills in a blank in the next sentence. A within-subject experiment i...

In the Story Cloze Test, a system is presented with a 4-sentence prompt to a story, and must determine which one of two potential endings is the ‘right’ ending to the story. Previous work has shown that ignoring the training set and training a model on the validation set can achieve high accuracy on this task due to stylistic differences between the story endings in the training set and validat...